The bioartificial liver (BAL) has been proposed as a support device during reversible hepatic injury or as a bridge technology in liver transplantation. The investigators have developed a BAL that provides support in an anhepatic rabbit model. In order to scale-up for human use, improved liver cell function and increased viable cell density are required. Our BAL is a three compartment hollow fiber bioreactor with two intraluminal compartments, one containing gel-entrapped rat hepatocytes and one containing media surrounding the hepatocyte gel; blood from a patient or animal in liver failure circulates in the (third) extracapillary compartment. Initial studies of the collagen- entrapped hepatocytes in static culture demonstrated more than 50% viability at eight days, cell aggregation with bile canaliculus formation, ureagenesis, amino acid uptake, albumin synthesis, and the ability to conjugate bile salts and metabolize lidocaine. When tested in acute anhepatic rabbit experiments, the BAL was not associated with clotting or hemolysis; it maintained normal glucose homeostasis and normalized plasma amino acids; urea levels were stabilized and lidocaine was biotransformed. No rabbit antibody reached the inner lumen gel- entrapped rat cells, a reflection of the bioreactor membrane pore size. The current BAL should support a 3-fold increase in viable cell density. We hypothesize that hepatocyte function and viable cell density can be improved through the use of extracellular matrix, growth factors, drug induction and cell-cell coculture; in addition, the enhanced BAL will demonstrate improved biochemical function in the anhepatic rabbit and prolonged survival in a D-galactosamine liver injury model. Preliminary data with the collagen-entrapped hepatocytes in static culture showed improved viability, albumin synthesis, bile acid conjugation and lidocaine metabolism when specific matrix factors or growth factors were added to the system. These observations will be extended in a modified linear optimization protocol to identify the best combination of effectors on the following markers of hepatocyte function: cell viability, cell aggregation, albumin synthesis, bile acid and bilirubin conjugation, and 4-methylumbelliferone and lidocaine metabolism. Specific effectors of hepatocyte viability and function include bovine intestinal heparin sulfate, pig liver glycosaminoglycan extract, Matrigel, recombinant hepatocyte growth factor, hormone enhanced media, lipocyte coculture, and drug induction with phenobarbital, clofibric acid, or beta-naphthoflavone. Effectors with the greatest influence on viability and function will be evaluated with the D-galactosamine hepatic injury model in a randomized, prospective study. This process should enable a substantial (50-70%) reduction in BAL scale-up that is necessary for larger animal and, eventually, human testing.